Biased gene expression in early honeybee larval development

Abstract

Background: Female larvae of the honeybee (Apis mellifera) develop into either queens or workers depending on nutrition. This nutritional stimulus triggers different developmental trajectories, resulting in adults that differ from each other in physiology, behaviour and life span.

Results: To understand how these trajectories are established we have generated a comprehensive atlas of gene expression throughout larval development. We found substantial differences in gene expression between worker and queen-destined larvae at 6 hours after hatching. Some of these early changes in gene expression are maintained throughout larval development, indicating that caste-specific developmental trajectories are established much earlier than previously thought. Within our gene expression data we identified processes that potentially underlie caste differentiation. Queen-destined larvae have higher expression of genes involved in transcription, translation and protein folding early in development with a later switch to genes involved in energy generation. Using RNA interference, we were able to demonstrate that one of these genes, hexamerin 70b, has a role in caste differentiation. Both queen and worker developmental trajectories are associated with the expression of genes that have alternative splice variants, although only a single variant of a gene tends to be differentially expressed in a given caste.

Conclusions: Our data, based on the biases in gene expression early in development together with published data, supports the idea that caste development in the honeybee consists of two phases; an initial biased phase of development, where larvae can still switch to the other caste by differential feeding, followed by commitment to a particular developmental trajectory.

Figure 1

Model depicting the major findings from this study. In this study we have identified substantial differences in gene expression that can be detected as early as 6 hours after exposure to RJ. As indicated by the arrows, a small number of these genes maintain higher expression throughout larval development in either the queen caste (red arrow, mitochondrial cytochrome C, phosphoenolpyruvate carboxykinase (PEPCK), phytanoyl-CoA dioxygenase domain-containing protein 1 homolog and glycine N-methyltransferase-like (GNMT)) or worker caste (blue arrow, Bcl-2). The differences in gene expression occur earlier in larval development than previously reported [8,15] and before the point at which we know larvae become committed to a particular caste [6]. These early, and sustained, differences in gene expression have led us to propose that this phase of early development represents a biased developmental trajectory. During this phase of early larval development gene expression is altered in response to RJ, biasing towards queen development (red line), or worker development (blue line) but not irreversibly so. Following this period, corresponding with peak JH levels in queens, we observe unique sets of genes induced in queen (red arrow, juvenile hormone-inducible (Jhl-26)) and worker castes (blue arrow, translocator protein, shep-like, lethal(2) essential for life and LOC10058039) that remain stably more highly expressed throughout the remainder of larval development. We propose that these changes in gene expression at 84 hours of larval development are associated with the induction of a committed developmental trajectory.

Figure 2

The CYP gene family is extensively differentially regulated during caste development. CYP genes are listed below the caste and time-point where they are differentially expressed. For clarity only CYP genes that are differentially expressed are shown. Genes are colour coded according to the clan they are assigned to based on based on sequence similarity [38]. The total number of CYP genes identified in the honeybee genome for each of the four clans [38] is indicated beside the colour key.

Figure 3

Hexamerin 70b expression biases caste determination towards worker fate. A. Relative expression of the four hexamerin genes in 60 hour queen and worker larvae as determined by RT-qPCR. Error bars represent the standard error of the mean. Statistical significance was assessed using a one-way ANOVA, *** indicates P < 0.001. B. Relative expression of hex70b was determined 12 – 18 hours following injection of either hex70b or eGFP (control) dsRNA using RT-qPCR. Error bars represent the standard error of the mean. Statistical significance was assessed using a one tailed t-test. C. The phenotype of emerging bees was assessed using a suite of morphological characteristics (refer to Additional file 1 for more detail). Graph shows the percentage of individuals that developed as queen-like after RNAi against hex 70b or eGFP (control). Statistical significance was assessed using a t-test, * indicates P < 0.05.

Figure 4

Annotation of genes involved in the spliceosome showing genes that are more highly expressed in queen larvae. The left hand side of the figure depicts the assembly of the spliceosome on a primary mRNA. Genes encoding proteins involved in the spliceosome are depicted as oblong boxes, with those more highly expressed in queen larvae shown in red and the remaining genes in grey (no genes were more highly expressed in worker larvae). The five small nuclear ribonucleoproteins that make up the spliceosome are depicted as circles (U1-U5). The spliceosome associated complexes Prp19 and EJC/TREX and are also shown.

Figure 5

Association of differentially expressed genes with alternative splicing. (A) Percentage of genes encoding single (dark bars) and multiple transcripts (light bars) as determined by HTS at 60 hours of larval development. Statistical significance was assessed using a Chi-squared test, * P < 0.05, ** P < 0.01. The number of genes in each category is indicated on the graph. (B). Percentage of DEGs where a single transcript (dark bars) or multiple transcripts (light bars) are differentially regulated. The number of genes in each category is indicated on the graph.